Sold by: Kooditurva Oy
Deployed on AWS
Deep-learning survival analysis (DeepSurv) on SageMaker: predict time-to-event and remaining useful life from censored data. Free tier: training on up to 100 units.
Overview
Lifespan Predictor is a SageMaker training algorithm for survival analysis, also called time-to-event analysis. It learns from historical records where some events have been observed and others haven't happened yet (censored). It then predicts how long until the event happens for new subjects, as a full survival curve rather than a single number.
Lifespan Predictor trains DeepSurv, a feed-forward neural network that generalises the Cox proportional hazards model. A classical Cox model assumes each feature affects risk linearly. DeepSurv learns non-linear effects and interactions between features from the data, while keeping the Cox model's handling of censored observations. The model is built on PyTorch.
Preprocessing included:
- automated feature engineering and scaling;
- optional rolling-window features (for example, rolling means over 3, 5 and 7 time steps) to summarise sensor history;
- per-unit handling of training, validation and test data, so records from one machine or customer never leak between sets.
- automated hyper parameter optimisation
Main use cases:
- Predictive maintenance: remaining useful life of equipment from sensor readings.
- Customer churn: time until a customer leaves.
- Conversion timing: time until purchase or activation.
This is a free tier version where training is capped to max 100 units. For access to paid version, contact info@kooditurva.fi .
Highlights
- Uses DeepSurv which is a neural extension of the Cox proportional hazards model that learns non-linear feature effects.
- Does automated feature engineering and automated hyper parameter optimisation which you control via training parameter dictionary.
- Handles censored data and returns full survival curves, not only point estimates.
Details
Sold by
Categories
Delivery method
Latest version
Deployed on AWS
New
Introducing multi-product solutions
You can now purchase comprehensive solutions tailored to use cases and industries.
Features and programs
Financing for AWS Marketplace purchases
AWS Marketplace now accepts line of credit payments through the PNC Vendor Finance program. This program is available to select AWS customers in the US, excluding NV, NC, ND, TN, & VT.
Pricing
This product is available free of charge. Free subscriptions have no end date and may be canceled any time.
Additional AWS infrastructure costs may apply. Use the AWS Pricing Calculator to estimate your infrastructure costs.
Dimensions summary
This is a free listing, so none of the ten dimensions carry a software charge. They split across three activities on SageMaker: training, batch inference, and real-time inference. Each activity is further divided by the compute instance size you choose, from ml.m5.large up to ml.m5.4xlarge. All are billed by host-hour, meaning you pay only for the time the instance runs. Larger instance sizes offer more compute per hour. You pick the activity and instance size that match your workload. The free tier supports fleets of up to 100 units.
Top-of-mind questions for buyers
A host-hour is one running compute instance billed for one hour. Each instance you launch meters its own time. Training, batch inference, and real-time inference each count their instance-hours separately. Larger instance sizes still bill per hour, but give more compute during that hour.
A real-time endpoint bills for as long as it exists, even when it serves no predictions. Deleting the deployed model and its endpoint stops that time from accruing. Batch inference instead runs a transform job and meters only while that job runs.
Batch inference runs a transform job on an instance you pick, metering host-hours only while that job runs. Real-time inference hosts the model on a standing endpoint that meters host-hours continuously until deleted. Batch suits scheduled scoring; real-time suits on-demand requests.
kooditurva.fi+1
Helpful?
Vendor refund policy
This product is free of software charges, so there is nothing to refund. You pay AWS only for the SageMaker infrastructure you use.
Legal
Vendor terms and conditions
Upon subscribing to this product, you must acknowledge and agree to the terms and conditions outlined in the vendor's End User License Agreement (EULA) .
Content disclaimer
Vendors are responsible for their product descriptions and other product content. AWS does not warrant that vendors' product descriptions or other product content are accurate, complete, reliable, current, or error-free.
Delivery details
Amazon SageMaker algorithm
An Amazon SageMaker algorithm is a machine learning model that requires your training data to make predictions. Use the included training algorithm to generate your unique model artifact. Then deploy the model on Amazon SageMaker for real-time inference or batch processing. Amazon SageMaker is a fully managed platform for building, training, and deploying machine learning models at scale.
Deploy the model on Amazon SageMaker AI using the following options:
Version release notes
Adds an interpretable penalised Cox model (model_type='coxph') with global and per-unit analyses.
Additional details
Inputs
- Summary
CSV with a header row: the unit identifier column "unit", the time column and the feature columns used in training, in the same order, without the labels. One row per observation.
- Limitations for input type
- CSV only (text/csv). Columns must match those the model was trained on.
- Input MIME type
- text/csv
unit,cycles,setting_1,setting_2,setting_3,f1,f2,f3,f4,f5,f6,f7,f8,f9,f10,f11,f12,f13,f14,f15,f16,f17,f18,f19,f20,f21
1,1,0.0023,0.0003,100.0,518.67,643.02,1585.29,1398.21,14.62,21.61,553.9,2388.04,9050.17,1.3,47.2,521.72,2388.03,8125.55,8.4052,0.03,392,2388,100.0,38.86,23.3735
1,2,-0.0027,-0.0003,100.0,518.67,641.71,1588.45,1395.42,14.62,21.61,554.85,2388.01,9054.42,1.3,47.5,522.16,2388.06,8139.62,8.3803,0.03,393,2388,100.0,39.02,23.3916
1,3,0.0003,0.0001,100.0,518.67,642.46,1586.94,1401.34,14.62,21.61,554.11,2388.05,9056.96,1.3,47.5,521.97,2388.03,8130.1,8.4441,0.03,393,2388,100.0,39.08,23.4166
https://github.com/Kooditurva/lifespan-predictor-sdk/blob/v1.1.0/samples/batch-input.csv
Resources
Vendor resources
Support
Vendor support
Contact info@kooditurva.fi on any matters.
AWS infrastructure support
AWS Support is a one-on-one, fast-response support channel that is staffed 24x7x365 with experienced and technical support engineers. The service helps customers of all sizes and technical abilities to successfully utilize the products and features provided by Amazon Web Services.
Similar products
Prediction of Patient’s Length of Stay in Hospitals
Take the guesswork out of the bid process and open the door to new possibilities for successful bid creation with Intellicision’s Bid Predictor. Whether you’re preparing a single quote, multiple quotes, or looking for a price evaluation, the Predictor illuminates the most direct path to a winning bid.
What's the difficulty level of this article? Evaluate the difficulty level of a word, sentence, or paragraph and compare it with another.
Identifies people who are likely to have a heightened vulnerability to severe complications from COVID-19
Customer reviews
No customer reviews yet
Be the first to review this product . We've partnered with PeerSpot to gather customer feedback. You can share your experience by writing or recording a review, or scheduling a call with a PeerSpot analyst.